Pike-devel

pike-devel@lists.lysator.liu.se

3 participants
3048 discussions

Problem with function prototypes
by Alexander Demenshin 29 Feb '04

29 Feb '04

Hi, Doing some tests, I discovered interesting behavior... Say, we have the following: class Test { string method(int n, string s) { // ...blabla } } And some declarations: Test typed = Test(); object untyped = Test(); So now, when I try to call typed->method(), the prototype will be checked, but when I call untyped->method(), it won't (and all arguments will be 0). As I understand, this is exactly because of absense of type declaration in latter case, but... Is it intended to behave like this? Sure, run-time prototype checking is expensive, but... It's absense may lead to some strange application behavior, especially in C module (do I've to call pop_n_elems(args) regardless of function prototype?). After all, it places responsibility to check all arguments to the developer, explicitly (so it has to be done anyway). Concerning C modules... I have something like this: void object_match(INT32 args) { struct pike_string *s; // ... get_all_args("match", args, "%S", &s); // ... } And prototype defintion like this (in pike_module_init()): add_function("match", object_match, "function(string:int)", 0); So, when this method is accessed through generic typed variable (object re) without arguments, I get segfault in 7.5, while in 7.4 everything is OK (error is thrown with no [visible] side effects). I wonder what will happen if function will be called with arguments while it accepts none (and there is no pop_n_elems(args) call)... Do I do something wrong? Or this is a bug somewhere? Regards, /Al

6 28

CVS access
by Alexander Demenshin 28 Feb '04

28 Feb '04

Hi, If possible, I would like to have write access to Pike's CVS. Primary goal - fixes to obvious (and sometimes not), like recent problem with get_all_args(). Yes, I've read and agreed to the policies, and I promise not to do any significant (or potentially dangerous) changes without prior discussion on the list. Regards, /Al

2 2

More 7.4/7.5 differences (C modules)
by Alexander Demenshin 28 Feb '04

28 Feb '04

Hi, Just another difference... Before (7.4), just using add_function() in pike_module_init() was enough to make it constant (i.e. visible through indices() and accessible by Module.func()). In 7.5, however, it is not the case - so anything that is added with add_function() is not visible (and not accessible). Is it right - so add_function_constant() should be used (I just want to make minimum changes to existing modules, while migrating to 7.5). Regards, /Al

3 10

create_process unable to find executable after chroot [3613]
by Henrik Grubbstr�m (Lysator) ＠ Pike (-) developers forum 28 Feb '04

28 Feb '04

I can't reproduce the problem. / Henrik Grubbström (Lysator) Previous text: >2004-02-28 09:38: >Subject: create_process unable to find executable after chroot [3613] >-------------------------------------------------------------------- > > Product: Pike > Version: 7.5 > Component: Core > Reporter: astra(a)upb.de > URL: https://community.roxen.com/crunch/show_bug.cgi?id=3613 > >The following does not work. There is a test file with exec permissions >/root/test and removing the chroot line will make it work. >It also fails with any other combination of PATH, /root/test, cwd, etc. > >void main() >{ > mapping env = getenv(); > System.chroot("/root"); > env["PATH"] = "/:/root/"; > mapping options = ([ > "stdin":Stdio.stdin, > "stdout":Stdio.stdout, > "env": env, > ]); > Process.create_process( ({ "test" }), options ); >} > > > / Brevbäraren >

2 1

modules in C++?
by Alexander Demenshin 27 Feb '04

27 Feb '04

Hi, Is it safe (or possible) to write modules in C++ instead of C? I don't any C++-safe declarations (extern "C") in Pike headers, so it seems - not... Do I miss something or? Regards, /Al

2 4

Nettle
by Martin Nilsson (saturator) ＠ Pike (-) developers forum 26 Feb '04

26 Feb '04

I PUSHED THE BUTTON. Now the _Crypto module isn't used (though still built). That means that you'll need Nettle installed to use any parts of Pike that uses Crypto.

9 31

7.4.44
by Martin Nilsson (saturator) ＠ Pike (-) developers forum 26 Feb '04

26 Feb '04

Peter and I have been working hard with resolving some NT problems and have now managed to create a new NT dist with pretty much everything except GLUT. 7.4.44 source dist and windows dist are available in FTP. Release notes at http://pike.ida.liu.se/download/notes/7.4.44.xml We'll release it more publicly when we have binary dists for more platforms. Please help out with good binary dists.

3 2

Pike 7.5 - CVS installation problem
by David Gourdelier 25 Feb '04

25 Feb '04

Hello, When trying to install Pike 7.5 - CVS on FreeBSD 4.3, I got this error: %make install /home/david/Pike/7.5/src/pike -DNOT_INSTALLED -DPRECOMPILED_SEARCH_MORE -m/home/david/Pike/7.5/src/master.pike -x extract_autodoc -q --srcdir=/home/david/Pike/7.5/src --imgsrc=/home/david/Pike/7.5/refdoc/src_images --builddir=doc_build/src --imgdir=doc_build/images Bad argument 1. Unknown program: foreach(0) /home/david/Pike/7.5/lib/modules/Tools.pmod/Standalone.pmod/extract_autodoc.pike:105: Tools.Standalone.extract_autodoc()->recurse("/home/david/Pike/7.5/src/doc_ build/src/doc_build/src/doc_build"+[963],"/home/david/Pike/7.5/src/doc_bui ld/src/doc_build/src/doc_build"+[977],0,({"predef::","doc_build","src","do c_build","src","doc_build","src",,,135})) /home/david/Pike/7.5/lib/modules/Tools.pmod/Standalone.pmod/extract_autodoc.pike:119: Tools.Standalone.extract_autodoc()->recurse("/home/david/Pike/7.5/src/doc_ build/src/doc_build/src/doc_build"+[954],"/home/david/Pike/7.5/src/doc_bui ld/src/doc_build/src/doc_build"+[968],0,({"predef::","doc_build","src","do c_build","src","doc_build","src",,,134})) ... After I tried to build Pike without the autodoc and I get compilation errors when using the Crypto module: > indices(Crypto); /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/cast.pike:4: Warning: Crypto.cast is deprecated. Use Crypto.CAST instead. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/module.pmod:12:Index 'crypt_md5' not present in module 'Crypto'. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/module.pmod:16:Index 'crypt_md5' not present in module 'Crypto'. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/aes.pike:4: Warning: Crypto.aes is deprecated. Use Crypto.AES instead. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/idea.pike:4: Warning: Crypto.idea is deprecated. Use Crypto.IDEA instead. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/sha.pike:4: Warning: Crypto.sha is deprecated. Use Crypto.SHA1 instead. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/hmac.pike:4: Warning: Crypto.hmac is deprecated. Use Crypto.HMAC instead. /usr/local/pike/7.5.17/lib/modules/Standards.pmod/PKCS.pmod/Signature.pmod:12:Index 'Hash' not present in module 'Crypto'. /usr/local/pike/7.5.17/lib/modules/Standards.pmod/PKCS.pmod/Signature.pmod:12:Illegal program identifier (type: int). /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/HMAC.pike:39:Error looking up 'Signature' in module 'PKCS'. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/HMAC.pike:77:Class definition failed. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/hmac.pike:6:Error looking up 'HMAC' in module 'Crypto'. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/hmac.pike:6:Illegal program pointer. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/hmac.pike:17:Class definition failed. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/hmac.pike:6:Error looking up 'hmac' in module 'Crypto'. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/hmac.pike:6:Illegal program pointer. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod.0:-: Warning: Compilation failed: Compilation failed. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/rsa.pike:4: Warning: Crypto.rsa is deprecated. Use Crypto.RSA instead. /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/RSA.pike:221:Failed to index module 'Signature' with 'build_digestinfo' (module doesn't exist?) /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/RSA.pike:291:Expected constant, got something else (438) /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/RSA.pike:339:Expected constant, got something else (438) /usr/local/pike/7.5.17/lib/modules/Crypto.pmod/RSA.pike:349:Expected constant, got something else (438) /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/rsa.pike:6:Error looking up 'RSA' in module 'Crypto'. /usr/local/pike/7.5.17/lib/7.4/modules/Crypto.pmod/rsa.pike:6:Illegal program pointer. ... / David

6 9

Embedding Pike into C/C++ game
by Fredrik (Naranek) Hubinette (Real Build Master) ＠ Pike (-) developers forum 25 Feb '04

25 Feb '04

In my oppinion, pike would have to be a lot more modular to become popular in games. Basically, you would need to chop pike into pieces that can be linked individually. In my mind, those pieces would be: 1) The interpreter Capable or running pike-code, but provides only the most basic functions needed to manipulate data types. The interpreter does not provide a way to load code into memory. 2) The Compiler Capable of compiling pike-code to in-memory code 3) Object loader (could be more than one) Capable of loading pre-compiled pike code into memory. 4) Function library. (or libraries) Provides all the functions needed to make up a complete Pike. 5) Pike executable Uses all of the above libraries to run Pike programs. Apart from this, all identifiers that don't start with 'Pike' should be fixed to make Pike more embedding-friendly. / Fredrik (Naranek) Hubinette (Real Build Master) Previous text: >2004-02-23 16:54: >Subject: Embedding Pike into C/C++ game >-------------------------------------------------------------------- >I forgot some things in my original email. This is mostly geared >towards the developers of Pike, but I'm copying it to the normal >mailing list as well. > >I wanted to point out that what Kenneth is saying probably is very >true. Pike is a very nice and powerful language. Python and Lua are >indeed the most commonly used languages in games today. However they >aren't without issues - Lua especially (for example very easy to make >mistakes that you don't notice until much later). Does Pike have a >place here? I think it's possible. > >It might have a future in games if it was easily embeddable (and >behaved nicely in those cases). Of course the other big hurdle is >documentation and lack of bugs. That is as we all know a problem >that's harder than embedding to solve. > >So the question is, how does this get solved? Is there a document or >notes about issues with embedding? For now, limited to embedding a >single interpreter in a process (not worrying about multiplicity as >Perl puts it, which is a whole other layer of complexity). Has anyone >sat down and looked at this more in detail? > > / David Hedbor >

4 4

XMLRPC patch for 7.2
by David Gourdelier 25 Feb '04

25 Feb '04

Hello, I have a patch for Pike 7.2 XMLRPC. The spec at http://www.xml-rpc.org/ says the <value> element can contain <string>, <array> and so on but can also contain a PCDATA by itself ("If no type is indicated, the type is string."). This wasn't set in XMLRPC.pmod and this is used by some client like XML-RPC Apache project one. Here is the patch: on Feb 23 16:38:40 2004 @@ -137,7 +137,7 @@ <!ELEMENT params (param*)> <!ELEMENT param (value)> - <!ELEMENT value (boolean|i4|int|double|string|base64| + <!ELEMENT value (#PCDATA|boolean|i4|int|double|string|base64| array|struct|dateTime.iso8601)> <!ELEMENT boolean (#PCDATA)> @@ -231,6 +231,8 @@ // seems to assume localtime). return Calendar.parse("%dT%h:%m:%s", data*"") || magic_zero; + // the spec says that the default is string inside value + default: return data*""; } error("Unknown element %O.\n", name); case "error": / David

3 5

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pike-devel